Naturalized Platonism
B. Linsky,E. Zalta
Abstract:objects are justified as a whole as part of our best scientific theories. We simply don’t need to justify our individual beliefs about particular abstract objects. (3) Armstrong and Maddy take Benacerraf’s problem seriously and respond by more thoroughly naturalizing the entities in question. Armstrong locates properties within the causal order, and Maddy does the same for sets. The Benacerraf problem then simply dissolves, at least in the context of a naturalized theory of truth and reference such as that described by Field. Since our primary objective in this paper is to put forward our own positive view, we shall not rehearse in detail our reasons for not adopting one of these responses. However, it is important for us to sketch what we take to be their most serious prima facie problems, if only for the purpose of contrast with our own view. Of course, many of the points we raise in the remainder of this section have appeared in the literature. We begin with Field’s view, even though he is not a Platonist, and so not a Naturalized Platonist. The main problems with his view are: 1. The complete dispensability of mathematics has not been established. It is doubtful whether the project can be carried out with respect to our most important physical theories, such as quantum mechanics and quantum field theory. 2. Even when mathematics is dispensable from actual science, it does not follow that it is dispensable from every scientific theory that might develop. We require an account of the language and subject matter of those portions of mathematics that might play a role in natural science, even if they don’t currently play a role. And we even 12“Epistemology and Nominalism,” published in A. Irvine, ed., Physicalism in Mathematics (Dordrecht: Kluwer, 1990), pp. 1-16. 13Universals and Scientific Realism, op. cit . 14Realism in Mathematics (Oxford: Clarendon, 1990); and “The Roots of Contemporary Platonism,” Journal of Symbolic Logic, 54/4 (December 1989): 1121-1144. 15See H. Field, “Tarski’s Theory of Truth,” The Journal of Philosophy, LXIX (1972): 347-375. C. Swoyer and B. Mundy together constitute a variant of this Armstrong and Maddy camp—these philosophers deny the indispensability of sets but accept the indispensability of properties. They have a thoroughly naturalistic view of properties. See C. Swoyer, “The Metaphysics of Measurement,” op. cit., and B. Mundy, “The Metaphysics of Quantity,” Philosophical Studies, 51 (January 1987): 29-54. Bernard Linsky and Edward N. Zalta 6 require an account of the dispensable portions of mathematics, if only to describe, as part of the very explanation of dispensability, the relation between the languages of natural science and mathematics. 3. Field uses the framework of second-order logic to show the dispensability of mathematics from classical physics. But he must reject the classical semantics of second-order quantifiers (as ranging over sets or properties). So his logic is developed using modal notions, and it is unclear whether those modal notions provide an adequate foundation for logic. 4. Field started his project by denying that numbers exist. But, what exactly is their status? Why is the language of number theory meaningful if its terms denote nothing at all? Field draws an analogy with fiction, claiming that ‘2+2=4’ is true only in sense in which ‘Holmes is a detective’ is true. But if numbers are useful fictions, then what is a fiction? No account is offered. 5. Recently, Field has suggested that numbers are abstract objects that happen not to exist. He accepts that they exist at other possible worlds. Field may have been led to this position for the following reasons. To explain the dispensability of mathematics, he attempts to establish its conservativeness, i.e., that there are no logical consequences of scientific theories involving mathematical claims that aren’t already consequences of the nonmathematical portion of the theory. But recall that his notion of consequence is not the usual model-theoretic one, but rather modal. To figure out whether one claim follows from another, you have to consider a world in which the latter claim is true. So in order to talk about the consequences of scientific theories involving mathematical claims, one must consider worlds where the mathematical claims are true. In such worlds, the numbers exist. So Field is led to accept that numbers might have existed, but in fact don’t. Yet if numbers don’t in fact exist but might have, then what is the conception of contingently existing abstract objects that underlies this position? Why should abstract objects exist at some worlds and not at others? 16Realism, Mathematics, and Modality (Oxford: Blackwell, 1989). 17This point is the subject of B. Hale and C. Wright, “Nominalism and the Contingency of Abstract Objects,” The Journal of Philosophy, LXXXIX/3 (March 1992): 7 Naturalized Platonism vs. Platonized Naturalism On the other hand, let us assume that Burgess is correct and that the Benacerraf problem has no force against Quine’s limited kind of Platonism. Quine’s view still faces certain other prima facie obstacles, however. The more serious ones are: 1. There is no account of mathematics that is not applied in scientific theories. Such mathematics certainly might be applied, and even if it is never applied, it is expressed in a meaningful language. How do we account for the meaningfulness of that language? 2. The mathematical portion of a scientific theory does not seem to receive confirmation from the empirical consequences derivable from the theory as a whole. Sober points out that there is a core of mathematical principles common to all competing scientific hypotheses. Since this core group of mathematical principles are assumed in every competing theory, evidence for the theory as a whole confers no incremental confirmation on the purely mathematical portion. Simply put, the evidence neither increases nor decreases the likelihood of those mathematical principles, since they are part of every competing hypothesis. This suggests that mathematics is not continuous with scientific theory. 3. If the overall scientific theory fails, scientists don’t revise the mathematical portion but instead switch to a different mathematical theory. The revolutions in physics in the early part of this century were accompanied by appeal to previously unapplied mathematical theories of non-Euclidean geometries, not by revising Euclidean geometry. Even in those cases where the needs of physical theories spurred the development of new mathematics, those needs never 111-135, and their followup article “A Reductio Ad Surdum? Field on the Contingency of Mathematical Objects,” Mind 103/410 (April 1994): 169-184. See also B. Linsky and E. Zalta, “In Defense of the Simplest Quantified Modal Logic,” in J. Tomberlin, ed., Philosophical Perspectives 8: Logic and Language (Atascadero, CA: Ridgeview Press, 1994), pp. 431-458. In that paper, the present authors introduce contingently nonconcrete objects in order to give an “actualistic” interpretation of the simplest quantified modal logic (i.e., a logic that includes the Barcan formulas). But Field could not appeal to those objects to ground his conception, for numbers are necessarily, rather than contingently, nonconcrete. 18Indeed, even for the scientific portion of the theory, different pieces of evidence seem to bear on different parts of the theory. Confirmation doesn’t seem to be holistic. 19“Mathematics and Indispensability,” The Philosophical Review , 102 (1993): 3557. Bernard Linsky and Edward N. Zalta 8 altered the normal a priori procedures of mathematical justification by axiomatization, definition, and proof. This point also casts doubt on the continuity of mathematics with natural science. 4. The account of logic doesn’t fit the facts. With the exception of quantum logic, no empirical evidence has ever been adduced in the course of arguing for alternative logics. The proliferation of alternative logics are not revisions of classical logic forced by empirical theory. Quantum logics stand alone, rather than as the first of a series of logics revised to suit the needs of physics. 5. The other problem posed by Benacerraf, concerning the arbitrariness of reductions, still remains. And even if other mathematical entities could be reduced to sets in a nonarbitrary way, it doesn’t follow that they are just sets. Mathematicians who are not working on set theory do not take themselves to be studying sets. There is a strong intuition that every mathematical object is what it is and not some other (mathematical) thing. Finally, we consider those philosophers who meet Benacerraf’s challenge by more thoroughly naturalizing Platonic entities such as sets or properties (i.e., by locating them in the causal order). By accepting Quine’s limited Platonism, Maddy inherits all of the problems just described (except for the first part of the last problem). But she and Armstrong face further difficulties as well: 1. For Maddy, there seems to be no way to assess the rationality of arguments for the highly theoretical axioms of ZF, such as the large cardinal axioms. This is the very part of the discipline that mathematicians find most interesting. 2. While Maddy solves the Benacerraf problem of arbitrary reductions by identifying numbers with structural properties of sets, the cost is that she denies the logical intuition (and common sense view of practicing mathematicians) that numbers are (individual) objects. 3. For Armstrong’s sparse conception of properties and states of affairs, there is a problem of finding enough properties and states to account both for natural science and the mathematics it requires, without accepting uninstantiated properties. 9 Naturalized Platonism vs. Platonized Naturalism 4. The combinatorial account of possibility Armstrong develops appeals to fictional entities (such as possible states of affairs), which don’t seem to be part of the